1. Field of the Invention
The present invention relates to a pixel defect compensating apparatus and a method of compensating pixel defect, particularly, relates to an apparatus and a method of compensating pixel defect of a solid-state image sensor by means of signal processing in an image pickup apparatus utilizing a solid-state image sensor of all-pixel readout system.
2. Description of the Related Art
Generally, it is well known that a pixel is deteriorated by sectional defect of semiconductor such as crystal defect in a solid-state image sensor such as a CCD (Charge-Coupled Device), which is produced from semiconductor. Such pixel defect as always adding a certain bias voltage to an image output in response to an amount of incident light appears on a monitor screen as a white spot in high intensity if a signal of the pixel defect is processed as it is. The white spot is called as a pixel of white defect.
FIG. 1 is a diagram of depicting a method of compensating pixel defect by a pixel defect compensating apparatus, which compensates a pixel of white defect of an interline solid-state image sensor, hereinafter called an interline CCD (IT-CCD), of the prior art. An IT-CCD reads out pixel signals in the interline transfer system. CCD signals, which are outputted from an IT-CCD, are processed to eliminate noise components prior to compensating pixels of white defect and adjusted to a predetermined signal level, and then converted into digital signals.
CCD signals, which are obtained from 2 pixels in horizontal direction and converted into a digital configuration, are mixed and called a luminance signal. The luminance signal is written in plurality of line memories by each line and compensates a pixel of white defect. FIG. 1 is an example of compensating pixels of white defect included in luminance signals for 3 lines with providing line memories for 2 lines. Pixels of white defect in luminance signals for 3 lines are compensated by utilizing luminance signals for total 3 lines, which are composed of luminance signals for 2 lines read out from the line memories and a luminance signal for one line to be inputted succeeding to the 2 lines.
In FIG. 1, each 3 pixels continuing in horizontal direction are arranged in continuing 3 lines, that is, a first line L0, a second line L1, and a third line L2 respectively. In other words, 3 pixels xe2x80x9caxe2x80x9d through xe2x80x9ccxe2x80x9d are arranged in the line L0, and 3 pixels xe2x80x9cdxe2x80x9d through xe2x80x9cfxe2x80x9d are arranged in the line L1, and 3 pixels xe2x80x9cgxe2x80x9d through xe2x80x9cixe2x80x9d are arranged in the line L2.
According to the method of compensating pixel defect of the prior art, a pixel of white defect at the pixels xe2x80x9cbxe2x80x9d, xe2x80x9cexe2x80x9d, and xe2x80x9chxe2x80x9d shown in FIG. 1 is compensated by utilizing luminance signals of total 9 pixels forming a matrix of 3 pixels horizontally and 3 pixels vertically. In addition thereto, the compensation is applied to luminance signals for 3 lines together and not for each line so as to prevent a vertical aperture signal from a pixel of white defect, which may be included in the vertical aperture signal.
Compensation of a pixel of white defect in the pixel xe2x80x9cexe2x80x9d is depicted first. Luminance signals of the pixels xe2x80x9cdxe2x80x9d and xe2x80x9cfxe2x80x9d are compared with each other and a lower signal level of them is defined xe2x80x9cY1Lxe2x80x9d as a representative value. Luminance signals of 6 pixels including pixels xe2x80x9caxe2x80x9d through xe2x80x9ccxe2x80x9d and xe2x80x9cgxe2x80x9d through xe2x80x9cixe2x80x9d are compared with each other and a highest signal level of them is defined xe2x80x9cY1Hxe2x80x9d as a representative value. Then, with comparing the representative values xe2x80x9cY1Lxe2x80x9d and xe2x80x9cY1Hxe2x80x9d, a larger value of them is defined xe2x80x9cY1xe2x80x9d. In case that a value of subtracting the value xe2x80x9cY1xe2x80x9d from a luminance signal value of the pixel xe2x80x9cexe2x80x9d exceeds a predetermined threshold value, it is discriminated that a pixel of white defect is presented in the pixel xe2x80x9cexe2x80x9d.
In the case of being discriminated that a pixel of white defect is presented in the pixel xe2x80x9cexe2x80x9d as mentioned above, the pixel of white defect is compensated by replacing the luminance signal value of the pixel xe2x80x9cexe2x80x9d with the value xe2x80x9cY1xe2x80x9d. On the other hand, in case that a value of subtracting the value xe2x80x9cY1xe2x80x9d from a luminance signal value of the pixel xe2x80x9cexe2x80x9d does not exceed a predetermined threshold value, it is discriminated that a pixel of white defect is not presented in the pixel xe2x80x9cexe2x80x9d. Then, the luminance signal value of the pixel xe2x80x9cexe2x80x9d is utilized as it is.
In case that a pixel of white defect in the pixel xe2x80x9cbxe2x80x9d is compensated, luminance signals of the pixels xe2x80x9caxe2x80x9d and xe2x80x9ccxe2x80x9d are compared with each other and a lower signal level of them is defined xe2x80x9cY0Lxe2x80x9d as a representative value. Luminance signals of pixels xe2x80x9cdxe2x80x9d through xe2x80x9cfxe2x80x9d are compared with each other and a highest signal level of them is defined xe2x80x9cY0Hxe2x80x9d as a representative value. Then, with comparing the representative values xe2x80x9cY0Lxe2x80x9d and xe2x80x9cY0Hxe2x80x9d, a larger value of them is defined xe2x80x9cY0xe2x80x9d. In case that a value of subtracting the value xe2x80x9cY0xe2x80x9d from a luminance signal value of the pixel xe2x80x9cbxe2x80x9d exceeds a predetermined threshold value, it is discriminated that a pixel of white defect is presented in the pixel xe2x80x9cbxe2x80x9d.
In the case of being discriminated that a pixel of white defect is presented in the pixel xe2x80x9cbxe2x80x9d as mentioned above, the pixel of white defect is compensated by replacing the luminance signal value of the pixel xe2x80x9cbxe2x80x9d with the value xe2x80x9cY0xe2x80x9d. On the other hand, in case that a value of subtracting the value xe2x80x9cY0xe2x80x9d from a luminance signal value of the pixel xe2x80x9cbxe2x80x9d does not exceed a predetermined threshold value, it is discriminated that a pixel of white defect is not presented in the pixel xe2x80x9cbxe2x80x9d. Then, the luminance signal value of the pixel xe2x80x9cbxe2x80x9d is utilized as it is.
In addition thereto, in case that a pixel of white defect in the pixel xe2x80x9chxe2x80x9d is compensated, luminance signals of the pixels xe2x80x9cgxe2x80x9d and xe2x80x9cixe2x80x9d are compared with each other and a lower signal level of them is defined xe2x80x9cY2Lxe2x80x9d as a representative value. Luminance signals of pixels xe2x80x9cdxe2x80x9d through xe2x80x9cfxe2x80x9d are compared with each other and a highest signal level of them is defined xe2x80x9cY2Hxe2x80x9d as a representative value.
Then, with comparing the representative values xe2x80x9cY2Lxe2x80x9d and xe2x80x9cY2Hxe2x80x9d, a larger value of them is defined xe2x80x9cY2xe2x80x9d. In case that a value of subtracting the value xe2x80x9cY2xe2x80x9d from a luminance signal value of the pixel xe2x80x9chxe2x80x9d exceeds a predetermined threshold value, it is discriminated that a pixel of white defect is presented in the pixel xe2x80x9chxe2x80x9d.
In the case of being discriminated that a pixel of white defect is presented in the pixel xe2x80x9chxe2x80x9d as mentioned above, the pixel of white defect is compensated by replacing the luminance signal value of the pixel xe2x80x9chxe2x80x9d with the value xe2x80x9cY2xe2x80x9d. On the other hand, in case that a value of subtracting the value xe2x80x9cY2xe2x80x9d from a luminance signal value of the pixel xe2x80x9chxe2x80x9d does not exceed a predetermined threshold value, it is discriminated that a pixel of white defect is not presented in the pixel xe2x80x9chxe2x80x9d. Then, the luminance signal value of the pixel xe2x80x9chxe2x80x9d is utilized as it is.
Accordingly, in case that a luminance signal level of one pixel is extremely higher than those of adjacent pixels, the pixel is discriminated as a pixel of white defect and compensated as mentioned above.
However, with respect to a recent single-unit video camera-recorder utilizing a solid-state image sensor such as a CCD, a need for obtaining one blur-free still picture signal composed of moving picture signals including even and odd fields is increasing in conjunction with promoting higher image quality. In some cases, a progressive scan CCD, hereinafter called a PS-CCD, of all-pixel readout system is utilized for such a single-unit video camera-recorder not an IT-CCD, which reads out a pixel signal in the interline transfer system.
In case that a pixel of white defect in a CCD signal obtained from such a PS-CCD is compensated by a pixel defect compensating apparatus of the prior art with a luminance signal, which is obtained by mixing 2 pixels in horizontal direction, it causes a problem of deteriorating resolution. Further, in case that a pixel of white defect is compensated by a CCD signal, prior to mixing 2 pixels in horizontal direction, with comparing the CCD signal with adjacent pixels within such a simple compensating area as disclosed in the prior art, it causes another problem that a pixel of white defect can not be correctly discriminated or can not be properly compensated due to difference of signal levels of each color component in a CCD signal outputted from a solid-state image sensor.
Accordingly, to improve the above mentioned problems, there provided a pixel defect compensating apparatus, which comprises extracting means for extracting a pixel signal outputted form a solid-state image sensor of all-pixel readout system and composed of a same color component from an array of pixel signals indicating one color component out of a plurality of color components of pixel information of one pixel within a predetermined area of a picture, pixel defect discriminating means for comparing pixel information of a specific pixel in the pixel signal extracted by the extracting means with pixel information of other pixels than the specific pixel in the pixel signal and for discriminating whether or not pixel defect is presented in the specific pixel, and pixel defect compensating means for compensating pixel information of the specific pixel by using pixel information of pixels other than the specific pixel in the pixel signal and for outputting a compensated pixel information, in case that the specific pixel is discriminated by the pixel defect discriminating means such that a pixel defect is presented, and for outputting the pixel information of the specific pixel as it is, in case that the specific pixel is discriminated by the pixel defect discriminating means such that a pixel defect is not presented in the specific pixel. The pixel defect discriminating means discriminate that there is presented a pixel defect in the specific pixel, in case that a value of subtracting a maximum signal level of pixels other than the specific pixel from the signal level of the specific pixel exceeds a predetermined value. The pixel defect compensating means compensate pixel information of the specific pixel by replacing the value of signal level of the specific pixel with the value of the maximum signal level, in case of being discriminated that a pixel defect is presented in the specific pixel.
According to an aspect of the present invention, there provided a method of compensating pixel defect, which comprises a step of extracting a pixel signal outputted from a solid-state image sensor of all-pixel readout system and composed of a same color component from an array of pixel signals indicating one color component out of a plurality of color components of pixel information of one pixel within a predetermined area of a picture, a step of comparing pixel information of a specific pixel in the pixel signal extracted by the step of extracting with pixel information of other pixels than the specific pixel in the pixel signal, a step of discriminating whether or not pixel defect is presented in the specific pixel, a step of compensating pixel information of the specific pixel by using pixel information of pixels other than the specific pixel in the pixel signal, in case that the specific pixel is discriminated by the pixel defect discriminating means such that pixel defect is presented, a step of firstly outputting a compensated pixel information, in case of being discriminated that pixel defect is presented in the specific pixel, and a step of secondly outputting the pixel information of the specific pixel as it is, in case that the specific pixel is discriminated such that pixel defect is not presented in the specific pixel.
In addition thereto, the step of discriminating pixel defect further discriminates that there is presented a pixel defect in the specific pixel, in case that a value of subtracting a maximum signal level of pixels other than the specific pixel from a signal level of the specific pixel exceeds a predetermined value, and the step of compensating pixel information of the specific pixel further compensates pixel information of the specific pixel by replacing the value of signal level of the specific pixel with the value of the maximum signal level, in case of being discriminated that a pixel defect is presented in the specific pixel.